Trigger word: franchesca_garasaki (mind the underscore!), recommend using either of the versions at a weight of about 0.8.<h3 id="about-the-versions:">About the versions:</h3>I recommend trying both models and see which one suits you more.<ul><li>v0.5 was trained via PixAI LoRA trainer and is a bit overcooked. It will produce better results from time to time, but at the cost of parts of her appearance applying to other characters and objects (e.g. clothing) present in the image.</li><li>v1 was trained via the old Civitai Trainer and is moderate at replicating the core concept, but will feel different to v0.5 in terms of color palette and detailing, even when used with the same settings.</li></ul>

Files with SHA256: 308e87bd2d617d864f11831e04f1200044fe785c81e4843434a266590f681f0e